Statistically optimal image reconstruction for multi-detector intensity diffraction tomography

Abstract

Intensity diffraction tomography (I-DT) is an in-line holographic imaging method for reconstructing the three-dimensional complex refractive index distribution of a weakly scattering object. Because it circumvents the phase retrieval problem of diffraction tomography, I-DT reconstruction methods may benefit a range of imaging problems involving optical and coherent X-ray radiation. In this work, we investigate the use of statistically complementary data, provided by multiple (> 2) in-line intensity measurements, for effective suppression of image noise in I-DT. The noise properties of the reconstructed images are demonstrated to depend strongly on the specification of measurement geometry. The effects of experimental uncertainties on the performance of I-DT is investigated also. Computer-simulation studies that are representative of a tomographic microscopy implementation of I-DT are presented.